Safety standards & testing for family beds

Family sleep is often discussed as though every child, mattress and sleeping arrangement were the same.

They are not.

A newborn in a crib is not the same situation as a mobile three-year-old sleeping beside a parent. A sofa is not a firm mattress. Two separate mattresses pushed together are not the same construction as one continuous sleeping surface. And a full-size family mattress is not a crib mattress.

This page separates three things clearly:

EN 16890’s published scope is expressly about mattresses for cots and cribs.^[1]

The current US mandatory crib-mattress rule similarly applies to crib mattresses. ASTM F2933-25 became the Commission-accepted version under 16 CFR Part 1241 effective 3 May 2026, with CPSC modifications retained.^{[2] [3]}

The US wording can be confusing because it uses the term “full-size crib mattress.” In this context, “full-size” means a standard-size crib mattress — approximately 27¼ × 51⅝ inches (about 69 × 131 cm) — not a full-size adult mattress.^[2]

A 240, 300 or 400 cm family mattress is therefore not simply an oversized version of the product these mattress standards were written for.

The accredited laboratories we have contacted have told us they cannot issue ordinary, in-scope EN 16890 or ASTM F2933/16 CFR 1241 crib-mattress certification for our full-size family mattress.

Because some of the physical questions are still useful. For example:

Those are sensible engineering questions regardless of whether a mattress is 70 cm or 400 cm wide.

Two of them deserve a note on context. “Can components separate?” and “can gaps develop?” were written for cribs: products designed for a child sleeping alone, at times out of the caregiver’s sight and hearing, where an undetected gap can matter enormously. A shared family bed is a different situation. A caregiver is typically on the same surface, within arm’s reach, and a child in discomfort will usually make itself heard. That context changes how quickly a problem is likely to be noticed, though, as we explain in section 2.4, we still design to avoid gaps rather than rely on a sleeping adult noticing them.

Where a published test method can meaningfully be transferred to our product, we use it as a benchmark. Then we add tests for something crib-mattress standards are not designed around:

adults and children sharing one large, loaded mattress.

One point that can disappear from discussions about bed-sharing is very simple: how much room does everyone actually have?

The Danish Health Authority explicitly addresses this. Its current public guidance states that if parents choose to sleep with an infant in their own bed, the infant should have as much sleeping space as it would have in its own bed.^[5]

Its more detailed professional guidance says:

The Danish Health Authority’s general child-safety material repeats the same basic principle: enough room that the infant is not squeezed by the adults, firm closely fitted mattresses, and preferably one large mattress.^[13]

That does not establish a clinical equation such as “a 300 cm bed reduces SIDS by X% compared with a 180 cm bed.” No such statistic has been established.

But it does establish that space, firmness, mattress joins and adult-induced deformation are relevant considerations.

“Baby”, “child” and “bed-sharing” are frequently discussed as if the risk were constant. It is not. US NIH data show:^[7]

But neither should information about a six-week-old newborn automatically be presented as though it applies identically to a mobile ten-month-old, a two-year-old, a four-year-old or an older child.

The American five-month threshold sometimes discussed online is another thing entirely. CPSC defines an “infant sleep product” as a product marketed or intended to provide sleeping accommodation for an infant up to five months old, unless another mandatory infant-sleep rule already covers it. CPSC also states that describing sleep as “attended” or “supervised” does not by itself take an infant sleep product outside the rule.^[4]

“Bed-sharing” can describe physically very different situations:

Those situations are not physically identical.

Japan is an interesting real-world example

A 2026 JAMA Network Open study notes that previous Japanese studies reported 44.9–76.2% bed-sharing across infancy. Researchers then examined 70,353 live births in Shiga Prefecture from 2018–2024 using detailed Child Death Review data; all SUID cases in the study underwent forensic autopsy.^[8]

What does this prove?

It does not prove that bed-sharing is risk-free. It does not prove that bed-sharing reduces SIDS. And it is not a controlled Japan-versus-US experiment. The Japanese study itself identified fatal sleep-related incidents involving mechanisms such as suffocation/strangulation in bed.^[8]

But it establishes an important observation:

That is one reason we do not think “bed-sharing” should be treated as one single exposure with one single level of risk, independent of the surface, environment, people involved and the child’s age.

There are well-known co-sleeping educators with large audiences who focus on the practical aspects of bed-sharing.

They are not regulators, government health authorities, certification bodies or product-testing laboratories. We mention them because they are influential within the practical co-sleeping community and address subjects such as mattress firmness, flatness and recovery, bed positioning, gaps and entrapment, floor-bed setups and the selection of adult mattresses.

Many share practical checks that families can carry out at home, along with recommendations on mattress firmness, bed positioning, fall prevention and how to manage gaps. Some also provide country-specific recommendations for ordinary adult mattresses.

These recommendations include mattresses from a range of manufacturers. They are typically assessed according to characteristics such as firmness, flatness, construction and materials, not because the adult mattresses are certified to EN 16890 or ASTM F2933.

No full-size adult mattress can be certified within the crib-mattress scope of these standards. This also applies to the mattresses that co-sleeping educators recommend online.

Adult mattresses may nevertheless hold legitimate certifications relating to areas such as organic production, substances or emissions. Such certifications can be extensive and relevant, but each covers its own defined scope.

That is exactly the distinction we want to preserve:

FamBed is not affiliated with the educators or mattress manufacturers referred to. No endorsement, partnership or other association is implied.

The Danish Health Authority already establishes the first principle: an infant sharing the parents’ bed should have enough sleeping space.^{[5] [6]}

Our reasoning from there is straightforward. If two adults and one or more children are going to occupy the same sleeping surface, we believe it makes more sense to give them more usable space rather than less. A 300 or 400 cm sleeping surface can physically provide:

We are not claiming “a 400 cm mattress reduces SIDS by X%.” There is no clinical evidence establishing such a number. Our conclusion is simpler:

Many families create more sleeping space themselves. Common solutions include:

We understand why families do it — and we do not claim every homemade setup is unsafe. But every new connection introduces another variable:

movement · height difference · separation · compression · gap

Our approach is different:

Our family bed is modular underneath because that makes a very large bed practical to manufacture, transport and install. But the final sleeping surface is designed so the sleeper is not lying directly on exposed mattress joins. One of the main components that makes that possible is our full-width topper.

The word “topper” sometimes creates particular concern in the United States, where separate European-style toppers are less central to normal mattress construction. Our topper is not a loose blanket, duvet, pillow or small pad underneath a child.

Our topper is a full-width structural comfort layer covering the complete family sleeping surface — above the underlying modules, spanning the complete width. It serves several practical functions.

One continuous upper surface

The supporting mattresses underneath can be modular. The sleeper lies on one topper spanning the complete width, so the normal sleeping surface does not expose the joins between the underlying mattress modules.

A component we can test

Rather than asking customers to assume that a topper behaves safely because of its name or material, we test physical questions directly. We test whether it can:

Those are measurable questions — see the Test Library below.

Easier hygiene and maintenance

Family beds are used by children. Spills, sweat, illness and toilet accidents are normal parts of family life. A removable topper system with a removable, washable cover, used together with appropriate mattress protection, makes it substantially easier to clean and maintain the upper sleeping surface without replacing the complete structural mattress system underneath.

We do not claim that a topper makes a bed “hygienic” by itself. Our point is practical: a serviceable, removable upper layer makes cleaning, washing and replacement easier than contaminating the entire structural mattress.

In our system, the topper is therefore not an afterthought. It is one of the components that turns transportable support modules underneath into one continuous and maintainable sleeping surface above.

Entrapment deserves serious attention — particularly for very young, non-mobile infants. Official safe-sleep guidance warns against an infant becoming trapped between an adult mattress and a wall, and crib/bed-rail standards devote extensive attention to openings and gaps.^{[15] [16]}

We think it is also important to recognise that the context is different. A crib is designed to contain an infant during independent sleep while the caregiver may be asleep or elsewhere. A parent sharing the same sleeping surface is physically beside the child. That proximity can mean the caregiver may notice movement, noise or a changed position sooner than if they are in another room.

But we do not treat a sleeping parent as a safety device:

Age also changes the mechanism. A non-mobile infant who cannot reliably reposition its head or body is fundamentally different from an independently mobile toddler. CPSC’s children’s portable bed-rail framework itself defines those rails as products for children who can get in and out of an adult bed unassisted, typically around 2–5 years old. The same rules still contain entrapment requirements, showing that mobility changes the context but does not make every trapping gap irrelevant.^[16]

Our engineering view is therefore:

A crib-mattress firmness test asks an important question: is the mattress surface excessively soft? We test that. But a family bed creates another obvious engineering question:

What happens to the smaller sleeper’s area when a 70, 90 or 120 kg adult lies nearby?

An adult can create:

The Danish Health Authority specifically points out that a heavier adult can create a mattress depression towards which an infant may roll.^[6]

So our test programme does not stop with an unloaded firmness test. We also measure the surface under representative adjacent adult loads. That is not an EN or ASTM certification test. It is a family-bed-specific engineering test because family beds create a loading condition that a crib-mattress test was not designed to evaluate.

For each published test we provide:

A note on the tests marked as scheduled: most of them formalise checks we have already performed in some form during product development, as part of our own practical engineering of the bed. What they have not yet been is documented under the current protocol, developed together with experienced test laboratories, and confirmed replicable by an external party. Until that standard of evidence exists, we list them as scheduled rather than passed.

A PASS means: the tested product configuration met the predetermined acceptance criterion documented for that particular test. It does not mean:

For our own protocols, we distinguish between:

Our internal Test Library is intended to make our engineering work more transparent, not to create our own regulatory system. Where a test is carried out internally, we label it INTERNAL ENGINEERING TEST. Where we use a method from an existing standard, we label it METHOD DERIVED FROM EN 16890 or ASTM/CPSC-DERIVED METHOD.

We do not use:

Every test is linked to an explicit product revision (e.g. CS-MATTRESS-2026-08-A). If a safety-relevant property changes — foam specification, spring system, topper core or thickness, quilt, cover construction, attachment system or module geometry — we record whether existing results remain representative or the test must be repeated.

We do not think the subject benefits from either extreme — “bed-sharing is always safe” or “every form of bed-sharing carries the same risk.” The physical reality is more complicated.

The details matter.

That is the standard we believe a purpose-built family-bed manufacturer should be held to.